CN112381805A

CN112381805A - Medical image processing method

Info

Publication number: CN112381805A
Application number: CN202011289961.0A
Authority: CN
Inventors: 石磊; 华铱炜; 柏慧屏; 杨忠程; 余沛玥
Original assignee: Hangzhou Yitu Medical Technology Co ltd
Current assignee: Hangzhou Yitu Medical Technology Co ltd
Priority date: 2020-11-17
Filing date: 2020-11-17
Publication date: 2021-02-19
Anticipated expiration: 2040-11-17
Also published as: CN112381805B

Abstract

The invention discloses a medical image processing method, which comprises the following steps: rib points in the 3D medical image are acquired. Determining a center point (x)₀，y₀)，x₀Associated with the rib point x coordinate, y₀Associated with the rib point y coordinate. Acquiring vertebra information in any frame of 2D medical image, wherein the vertebra information comprises: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, the Z coordinate of the vertebra point, the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is a preset distance away from the first line segment, and the first line segment passes through the frame of 2D medical image vertebra detectionThe frame is centered and parallel to the X-axis. And mapping the vertebra information in the multi-frame 2D medical images to a first coordinate system to obtain the medical images corresponding to the multi-frame 2D medical images. The scheme of the invention improves the mapping efficiency and the mapping accuracy, is also beneficial to the reading and diagnosis of doctors, and further improves the diagnosis efficiency and the diagnosis accuracy of the doctors to a certain extent.

Description

Medical image processing method

Technical Field

The invention relates to the technical field of medical treatment, in particular to a medical image processing method.

Background

Currently, X-ray photography equipment, Computed Tomography (CT) equipment, Nuclear Magnetic Resonance Imaging (NMRI) equipment, and the like can acquire cross-sectional images (which may also be referred to as 2D images) of various parts of a human body, and can also generate 3D images of various parts of the human body.

At present, the cross-sectional image acquired by the medical imaging device usually cannot correspond to the panoramic image, and therefore, the tissue in the cross-sectional image needs to be determined manually and corresponds to the tissue in the panoramic image. Taking the acquired chest image as an example, the chest image includes vertebrae, and after a frame of chest cross-sectional image is acquired, a doctor can determine which vertebrae in the panoramic image correspond to vertebrae in the frame of cross-sectional image according to experience, so as to determine a subsequent treatment plan. However, in this way, due to different experiences of different doctors, the determined mapping accuracy of the cross-sectional image and the vertebral panorama is greatly influenced by artificial subjective factors, and the artificial image viewing usually takes a long time, resulting in low working efficiency and diagnosis efficiency.

Therefore, a medical image processing method is needed to solve the technical problems of the prior art that the mapping accuracy is greatly affected by artificial subjective factors and the efficiency is low due to the fact that a 2D medical image is mapped into a vertebral panorama by an artificial method.

Disclosure of Invention

The invention provides a medical image processing method, a medical image processing device and computer equipment, and aims to solve the technical problems that in the prior art, the mapping precision is greatly influenced by artificial subjective factors and the efficiency is low because a 2D medical image is mapped into a vertebral panoramic image in an artificial mode.

The invention provides a medical image processing method, which comprises the following steps:

acquiring rib points in the 3D medical image;

determining a center point (x)₀，y₀)，x₀Associated with the x-coordinate, y, of the rib point₀A Y-coordinate associated with the rib point, the X-axis being perpendicular to the sagittal plane and the Y-axis being perpendicular to the coronal plane;

acquiring vertebra information in any frame of 2D medical image, wherein the vertebra information comprises: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, wherein the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is away from the first line segment by a preset distance, and the first line segment passes through the center of the vertebra detection frame of the frame of 2D medical image and is parallel to the X axis;

mapping vertebra information in the multiple frames of 2D medical images to a first coordinate system to obtain medical images corresponding to the multiple frames of 2D medical images.

Optionally, the medical image processing method further includes:

rib information in any frame of 2D medical image is acquired, wherein the rib information comprises: the center point and the direction of a straight line where each rib point in the frame of 2D medical image is located, and the Z coordinate of the rib point;

and mapping rib information in the multi-frame 2D medical image to a first coordinate system.

Optionally, the rib points include: rib positioning point, associated positioning point associated with said rib positioning point, said x₀X-coordinate, y, associated with the positioning point of the rib₀The y-coordinate associated with the rib positioning point.

Optionally, the rib points include: rib positioning point, associated positioning point associated with the rib positioning point, rib contour point, the x₀X-coordinate, y, associated with the positioning point of the rib₀The y-coordinate associated with the rib positioning point.

Optionally, the mapping vertebral information in the multiple frames of 2D medical images to the first coordinate system includes:

for vertebra information in any frame of 2D medical image, mapping the direction of a straight line where a central point and a vertebra point are located to a first coordinate axis, and mapping the Z coordinate of the vertebra point to a second coordinate axis;

the mapping rib information in the multi-frame 2D medical image to a first coordinate system comprises:

for rib information in any frame of 2D medical image, mapping the direction of a straight line where a center point and a rib point are located to a first coordinate axis, and mapping the Z coordinate of the rib point to a second coordinate axis;

wherein the first coordinate axis and the second coordinate axis are perpendicular.

Optionally, the medical image processing method further includes acquiring a rib positioning point, where the acquiring a rib positioning point includes:

acquiring a first positioning point and a second positioning point in the 3D medical image, wherein the first positioning point is positioned in the left lung, and the second positioning point is positioned in the right lung;

determining a first axis and a second axis, wherein the first axis passes through the first positioning point and is perpendicular to the cross section, and the second axis passes through the second positioning point and is perpendicular to the cross section;

segmenting the 3D medical image by a section passing through the first axis and taking the first axis as a boundary so as to obtain a plurality of first tangent planes;

cutting the three-dimensional chest image by using a section which passes through the second axis and takes the second axis as a boundary so as to obtain a plurality of second cutting planes;

the plurality of first tangent planes and the plurality of second tangent planes are detected to obtain rib positioning points.

Optionally, the first positioning point is a central point or a center of gravity of the left lung, and the second positioning point is a central point or a center of gravity of the right lung.

Optionally, the length of the first line segment is equal to the length of a first edge of the vertebra detection frame, and the first edge of the vertebra detection frame is parallel to the X axis.

The present invention also provides a medical image processing apparatus, comprising:

a first acquisition unit for acquiring a rib point in the 3D medical image.

A determination unit for determining a center point (x)₀，y₀)，x₀Associated with the x-coordinate, y, of the rib point₀A Y-coordinate associated with the rib point, the X-axis being perpendicular to the sagittal plane and the Y-axis being perpendicular to the coronal plane;

a second obtaining unit, configured to obtain vertebra information in any one frame of the 2D medical image, where the vertebra information includes: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is a preset distance away from the first line segment, and the first line segment passes through the center of the vertebra detection frame of the frame of 2D medical image and is parallel to the X axis.

The first mapping unit is used for mapping vertebra information in the multi-frame 2D medical images to a first coordinate system so as to obtain the medical images corresponding to the multi-frame 2D medical images.

The invention also provides a computer device comprising at least one processor and at least one memory, wherein the memory stores a computer program which, when executed by the processor, enables the processor to carry out the medical image processing method as described above.

The present invention also provides a computer-readable storage medium, in which instructions, when executed by a processor in an apparatus, enable the apparatus to perform the above-described medical image processing method.

The invention also provides a medical image processing method, which comprises the following steps:

acquiring rib points in the 3D medical image;

acquiring a vertebra detection frame in any frame of 2D medical image;

translating the vertebra detection frame in each acquired frame of 2D medical image to obtain a translated vertebra detection frame corresponding to the vertebra detection frame, wherein Y coordinates of the center of the translated vertebra detection frame in each frame of 2D medical image are the same;

acquiring vertebra information in any frame of 2D medical image, wherein the vertebra information comprises: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, wherein the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is away from the first line segment by a preset distance, and the first line segment passes through the center of the translated vertebra detection frame of the frame of 2D medical image and is parallel to the X axis;

Optionally, the medical image processing method further includes:

the device comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring rib points in a 3D medical image;

the third acquisition unit is used for acquiring a vertebra detection frame in any frame of 2D medical image;

the translation unit is used for translating the vertebra detection frame in each acquired frame of 2D medical image to obtain a translated vertebra detection frame corresponding to the vertebra detection frame, wherein Y coordinates of the center of the translated vertebra detection frame in each frame of 2D medical image are the same;

a fourth obtaining unit, configured to obtain vertebra information in any frame of the 2D medical image, where the vertebra information includes: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, wherein the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is away from the first line segment by a preset distance, and the first line segment passes through the center of the translated vertebra detection frame of the frame of 2D medical image and is parallel to the X axis;

the second mapping unit is used for mapping the vertebra information in the multi-frame 2D medical images to the first coordinate system so as to obtain the medical images corresponding to the multi-frame 2D medical images.

Compared with the prior art, the technical scheme of the invention has the following beneficial effects:

by acquiring rib points in a 3D medical image and determining a center point based on the acquired rib points. Acquiring vertebra information in any frame of 2D medical image, wherein the vertebra information comprises: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is a preset distance away from the first line segment, and the first line segment passes through the center of the vertebra detection frame of the frame of 2D medical image and is parallel to the X axis. And finally mapping vertebra information in the multi-frame 2D medical images to a first coordinate system to obtain the medical images corresponding to the multi-frame 2D medical images. Because the multi-frame 2D medical images can be automatically mapped into the vertebra panorama without manually and subjectively judging the corresponding relation between the 2D medical images and the vertebra panorama, the mapping efficiency and the mapping accuracy are improved, and meanwhile, the reading and the diagnosis of a doctor are facilitated, and the diagnosis efficiency and the diagnosis accuracy of the doctor are further improved to a certain degree.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:

FIG. 1 is a schematic representation of the human body's basic faces and basic axes in a standard anatomy in accordance with an embodiment of the present invention;

fig. 2 is a flowchart illustrating a medical image processing method according to a first embodiment of the invention;

FIG. 3 is a schematic cross-sectional view of a 3D medical image being segmented according to an embodiment of the invention;

FIG. 4 is a schematic structural diagram of a first detection model according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a first tangential plane and rib locations detected thereon according to an embodiment of the present invention;

FIG. 6 is a schematic view of vertebral points according to an embodiment of the present invention;

fig. 7 is a flowchart illustrating a medical image processing method according to a second embodiment of the invention;

fig. 8 is a flowchart illustrating a medical image processing method according to a third embodiment of the present invention;

fig. 9 is a schematic diagram of key points of an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail with reference to the accompanying drawings, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As described in the prior art, currently, when a doctor reads a picture, the vertebra in a multi-frame sectional image is usually mapped into a vertebra panorama by an artificial method, and the mapping is performed by the artificial method, so that the mapping speed is low, and mapping errors are likely to occur, thereby causing the lower efficiency of the doctor in reading the picture and diagnosing the picture.

In order to better explain the technical scheme of the invention, the basic planes and the basic axes of the human body in the standard anatomy are explained correspondingly in the embodiment. Referring to fig. 1, fig. 1 is a schematic view of the basic planes and axes of a human body in a standard anatomy according to an embodiment of the present invention. As shown in fig. 1: the basic planes of the human body include the coronal plane (frontal plane), the sagittal plane (median plane), and the transverse plane (transverse plane, horizontal plane). The human body's basic axes include the vertical axis (the Z-axis, which is the axis from the top, down to the tail, and perpendicular to the ground plane), the sagittal axis (the Y-axis, which is the axis from the front to the back, and crossing the vertical axis at right angles), and the coronal axis (the X-axis, also called frontal axis, which is the axis parallel to the horizontal and perpendicular to the first two axes). The three-dimensional orientation of the human body includes a front side (close to the abdomen), a rear side (close to the back), an upper side (close to the head), and a lower side (close to the feet). All faces, axes, orientations in this embodiment apply to fig. 1.

Fig. 2 is a flowchart illustrating a medical image processing method according to a first embodiment of the invention. As shown in fig. 2, the medical image processing method of the present embodiment includes:

s11: rib points in the 3D medical image are acquired.

S12: determining a center point (x)₀，y₀)，x₀Associated with the x-coordinate, y, of the rib point₀The Y-coordinate associated with the rib points, the X-axis is perpendicular to the sagittal plane and the Y-axis is perpendicular to the coronal plane.

S13: acquiring vertebra information in any frame of 2D medical image, wherein the vertebra information comprises: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is a preset distance away from the first line segment, and the first line segment passes through the center of the vertebra detection frame of the frame of 2D medical image and is parallel to the X axis.

S14: mapping vertebra information in the multiple frames of 2D medical images to a first coordinate system to obtain medical images corresponding to the multiple frames of 2D medical images.

In step S11, in this embodiment, the 3D medical image may be a CT image including a chest image or an MR image including a chest image. The rib point is a point on the rib. Specifically, the rib point may be a point obtained by detecting a rib point in the 3D medical image, in the present case, the rib point obtained by detecting the 3D medical image is referred to as a rib positioning point, and the coordinate of the central point is associated with the coordinate of the rib positioning point.

In the embodiment, the 3D medical image is detected to obtain the rib positioning points through the following steps.

S111: and acquiring a first positioning point and a second positioning point in the 3D medical image, wherein the first positioning point is positioned in the left lung, and the second positioning point is positioned in the right lung.

S112: determining a first axis and a second axis, wherein the first axis passes through the first positioning point and is perpendicular to the cross section, and the second axis passes through the second positioning point and is perpendicular to the cross section.

S113: and cutting the 3D medical image by using a section which passes through the first axis and takes the first axis as a boundary so as to obtain a plurality of first tangent planes.

S114: and cutting the 3D medical image by using a section which passes through the second axis and takes the second axis as a boundary so as to obtain a plurality of second cutting planes.

S115: the plurality of first tangent planes and the plurality of second tangent planes are detected to obtain rib positioning points.

S111 is executed, in this embodiment, in order to obtain the first positioning point located in the left lung and the second positioning point located in the right lung, the 3D medical image needs to be segmented to obtain the left lung and the right lung. Specifically, a corresponding segmentation method may be selected according to actual applications, such as a thresholding method, a region growing method, a method based on pattern classification, a method based on image registration and a shape model, and the like, and the 3D medical image may also be segmented by a three-dimensional convolutional neural network model to obtain the lung image. This is not limited in this embodiment as long as the lung region can be segmented from the 3D medical image.

After the lung regions are segmented, a first localization point is determined in the left lung and a second localization point is determined in the right lung. In this embodiment, the first location point may be a center point or a center of gravity of the left lung, and the second location point may be a center point or a center of gravity of the right lung.

And S112, a first axis perpendicular to the cross section is made in the left lung through the first positioning point, and a second axis perpendicular to the cross section is made in the right lung through the second positioning point. The first axis runs through the left side of the 3D medical image and the second axis runs through the right side of the 3D medical image.

S113 is executed, in this embodiment, in order to avoid the influence of other tissues, vertebrae, and the like on the rib point detection, when the left side of the 3D medical image is sliced by the slice, the left side of the 3D medical image is sliced by using the first axis and any one plane with the first axis as a boundary as a slice to obtain a plurality of first slices. And the included angle between the intersection line of the first tangent plane and the cross section and the positive direction of the coronal axis (X axis) is more than or equal to-120 degrees and less than or equal to 120 degrees. That is, when the left side of the 3D medical image is sliced using the slice plane, the angle of the initial slice may be 120 ° and then the slices may be sliced at predetermined intervals in the counterclockwise direction until the slice is-120 °. Or when cutting, the angle of the initial cutting may be-120 ° and then the cutting is performed at predetermined intervals in the clockwise direction until the cutting reaches 120 °. Of course, in other embodiments, the initial slicing angle may be any angle between-120 ° and 120 °, as long as slicing between-120 ° and 120 ° is satisfied. In this embodiment, the predetermined interval may be less than or equal to 6 °, for example, the predetermined interval may be 3 °, and the cutting is performed every 3 ° by using the above-mentioned cutting plane to obtain a plurality of first cutting planes. In practical application, the predetermined interval may be determined according to actual requirements, and the smaller the predetermined interval is, the more rib positioning points are finally detected and obtained, and the higher the accuracy of the rib panorama finally obtained by mapping is.

S114 is performed to segment the right side of the 3D medical image. Similarly, in order to avoid the influence of other tissues, vertebrae, and the like on the detection of the rib point, when the right side of the 3D medical image is sliced by the slice, the right side of the 3D medical image is sliced by the second axis and any one plane with the second axis as a boundary as a slice to obtain a plurality of second sliced planes. And the included angle between the intersection line of the second tangent plane and the cross section and the negative direction of the coronal axis (X axis) is more than or equal to-120 degrees and less than or equal to 120 degrees. When the right side of the 3D medical image is sliced using the slice plane, the angle of the initial slice may be 120 °, -120 °, or any angle between-120 ° and 120 °. When the right side of the 3D medical image is sliced at an angle of 120 ° as the initial slice, slices are performed at predetermined intervals in the clockwise direction until the slice is sliced to-120 °. And when the right side of the 3D medical image is sliced at an angle of-120 ° as the initial slice, the slices are performed at predetermined intervals in the counterclockwise direction until the slice is sliced to 120 °. The predetermined interval may be 6 ° or less, for example, the predetermined interval may be 3 °, that is, the above-described cutting plane is used to perform cutting every 3 ° to obtain a plurality of second cutting planes. Similarly, in practical application, the predetermined interval may be determined according to actual requirements, and the smaller the predetermined interval is, the more the obtained rib positioning points are, thereby facilitating to improve the accuracy of the vertebra panoramic image obtained by final mapping.

Referring to FIG. 3, FIG. 3 is a schematic cross-sectional view of a 3D medical image according to an embodiment of the present invention, where P in FIG. 3₁Is the first site of orientation, P₂As the second anchor point, P is used in FIG. 3₁The rays as starting points are the projection of each first tangent plane on the cross section, with P₂The plurality of rays as the starting point are projections of the second tangent planes on the cross section. In FIG. 3, with P₁The included angle between the plurality of rays as the starting point and the positive direction of the coronal axis (X axis) is more than or equal to-120 degrees and less than or equal to 120 degrees in terms of P₂The included angle between the plurality of rays as the starting point and the negative direction of the coronal axis (X axis) is more than or equal to-120 degrees and less than or equal to 120 degrees.

S115 is executed to detect a plurality of first tangent planes to obtain rib positioning points of the left rib, and detect a plurality of second tangent planes to obtain rib positioning points of the right rib. In this embodiment, the first detection model may be used to detect the positioning point of the left rib on each first tangent plane. Specifically, the detection model may include: the device comprises a feature extraction module and a detection frame acquisition module. And the detection frame acquisition module detects the characteristic diagram output by the characteristic extraction module.

In this embodiment, the feature extraction module may include: l convolution units, M maximum pooling layers, N2 × 2D deconvolution layers, and a tensor superposition layer. Each convolution unit includes: a convolutional layer (Conv2d), a Batch Normalization layer (BN), and an activation layer, which may be a Linear rectification function (ReLU). In this embodiment, the feature extraction module may be a Feature Pyramid Network (FPN), and the detection frame acquisition module may be an ssd (single Shot multi box detector) network.

Referring to fig. 4, fig. 4 is a schematic structural diagram of a first detection model according to an embodiment of the present invention, in the embodiment, the first detection model includes: the feature extraction module and the detection frame acquisition module, wherein the feature extraction module includes: the first convolution unit to the eighth convolution unit, the first pooling layer to the third pooling layer, a 2X 2D deconvolution layer and a quantum superposition layer. In practical application, if the size of the first tangent plane to be detected is 512. 1, a tensor with size 64. 64 is output after passing through the first convolution unit, the second convolution unit, the first pooling layer, the third convolution unit, the fourth convolution unit, the second pooling layer, the fifth convolution unit, the sixth convolution unit, the third pooling layer in order, and the tensor with size 64. 32 continues to pass through the seventh convolution unit and the 2D deconvolution layer of 2. 32, resulting in a reconv 1 with size 128. 32. The tensor output by the second pooling layer passes through the eighth convolution unit to output a tensor conv8 with size 128 at 128, 32, and the tensor conv8 and the tensor deconv1 are added through a tensor addition layer to obtain a tensor add1 with size 128 × 32. And inputting the tensor add1 and the tensor conv7 output by the seventh convolution unit as two feature maps to a detection box acquisition module to detect rib positioning points on the first tangent plane. Specifically, a plurality of detection frames can be detected on two feature maps with different sizes by fixing an anchor in the SSD, the detection frame with the confidence higher than the preset threshold is retained (the preset threshold is determined according to actual requirements), then the detection frame with the higher overlap is removed by a Non-maximum suppression (NMS) method, and the finally obtained detection frame is the detection frame including the left rib positioning point detected on the first tangential plane. In this embodiment, the detection frame may be rectangular, and the center of the detection frame may be used as the positioning point of the left rib.

In this embodiment, the first detection model may use the multiple first tangent plane images marked with the rib positioning points as training samples, and may also perform an enhancement operation on the multiple first tangent plane images marked with the rib positioning points to expand the data volume of the training samples, where the enhancement operation includes: the image processing method includes the steps of randomly translating up, down, left and right predetermined pixels (for example, 0-30 pixels), randomly rotating a set angle (for example, -15 degrees), randomly zooming a set multiple (for example, 0.85-1.15 times), and slightly shaking the contrast and brightness of a first tangent plane image.

Inputting a training sample into an initial first detection model, calculating a loss function according to the marked rib positioning point and the rib positioning point detected by the initial first detection model during training, and repeatedly iterating by adopting a back propagation algorithm and a Stochastic Gradient Descent (SGD) optimization algorithm to continuously update parameters of the initial first detection model. If the loss function of a certain training is less than or equal to the threshold, the initial first detection model corresponding to the model parameter of the training can be used as the first detection model. After the first detection model is obtained through training, a first tangent plane obtained by segmenting the 3D medical image may be input to the first detection model to obtain a left rib positioning point on the first tangent plane. Referring to fig. 5, fig. 5 is a schematic diagram of a first tangential plane and the detected rib positioning points thereon according to the embodiment of the present invention, and fig. 5 shows three first tangential planes, and the detected left rib positioning points are shown on the first tangential plane.

Similarly, in this embodiment, a second detection model may be used to detect the right rib positioning point on each second tangential plane, and the structure of the second detection model in this embodiment is similar to that of the first detection model, except that when training the second detection model, a plurality of second tangential plane images marked with the rib positioning points are used as training samples. After the second detection model is obtained through training, the right rib positioning point on the second tangent plane can be obtained by inputting the second tangent plane obtained by segmenting the 3D medical image into the second detection model. In this embodiment, the left rib positioning point is detected through the first detection model and the right rib positioning point is detected through the second detection model in a targeted manner, so that the accuracy of detecting the left rib positioning point and the right rib positioning point is improved to a certain extent.

In other embodiments, only one detection model may be used to detect rib positioning points on the first tangent plane and the second tangent plane, and the structure of the detection model is similar to that of the first detection model, except that when training the detection model, the plurality of first tangent plane images and the plurality of second tangent plane images marked with the rib positioning points are used as training samples.

Thus, the rib points in the 3D medical image, i.e., the rib positioning points in the present embodiment, are detected through the above-mentioned steps S111 to S115. In this embodiment, the central point (x) may be obtained by the detected rib point (rib positioning point)₀，y₀) The position of (a).

Execution S12, determine center Point (x)₀，y₀). In this embodiment, specifically, the x coordinate of the central point may be an average value of the detected x coordinates of all rib positioning points (the sum of the x coordinate values of all rib positioning points is divided by the number of the rib positioning points), that is, the x coordinate₀Taking the average value of the detected y coordinates of all the rib positioning points (the sum of the y coordinate values of all the rib positioning points is divided by the number of the rib positioning points) as the y coordinate of the central point, namely the y coordinate₀。

Executing S13, obtaining vertebra information in each frame of 2D medical image, the vertebra information including: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is a preset distance away from the first line segment, and the first line segment passes through the center of the vertebra detection frame of the frame of 2D medical image and is parallel to the X axis. In this embodiment, in order to obtain the vertebra information in each frame of 2D medical image, it is necessary to first detect the vertebra in the frame of 2D medical image to obtain a vertebra detection frame.

In this embodiment, a vertebra detection model may be used to detect the position of the vertebra in the cross-section. The structure of the vertebra detection model is similar to that of the first detection model, except that in this embodiment, instead of directly inputting the frame of 2D medical image to be detected to the vertebra detection model, the 2D medical image of the frame to be detected is used as an intermediate frame image to obtain at least one frame of image before and after the frame of 2D medical image, and an image layer composed of the intermediate frame image and the at least one frame of image before and after the intermediate frame image is input to the vertebra detection model. Similarly, when the initial vertebra detection model is trained, an image layer composed of at least three images marked with vertebra positions in the intermediate frame images and at least one image positioned before and after the intermediate frame images is used as a training sample to train the initial vertebra detection model so as to obtain the vertebra detection model. For example, if the intermediate frame image is the 10 th frame image, at least one frame before the intermediate frame image may be the 9 th frame image, and at least one frame after the intermediate frame image may be the 11 th frame image. The at least two frames located before may be a 9 th frame image and a 8 th frame image, and the at least two frames located after may be an 11 th frame image and a 12 th frame image. In this embodiment, the image layer of the 3-frame image composed of the intermediate frame image and the frames before and after the intermediate frame image is input to the vertebra detection model to detect the vertebra in the intermediate frame image, for example, the 9 th frame image, the 10 th frame image and the 11 th frame image are input to the vertebra detection model to detect the position of the vertebra in the 10 th frame image. In this embodiment, the position of the vertebra is identified by means of a vertebra detection frame, the detection frame may be rectangular, and the center of the vertebra detection frame may be used as the position of the detected vertebra.

In this embodiment, the image layer is input to the vertebra detection model, so that the position information of the vertebra in the output intermediate frame image is combined with the position information of the vertebra in the previous frame and the vertebra in the next frame, and the detection accuracy of the vertebra detection frame in each frame of 2D medical image can be further improved.

After the vertebra detection frame in each frame of 2D medical image is obtained in the above manner, the vertebra information in the frame of 2D medical image is determined next. First, a vertebral point is determined. In this embodiment, the vertebra point is located on a first line segment passing through the center of the vertebra detection frame and parallel to the X axis, and in order to improve the accuracy of vertebra mapping, in this embodiment, the length of the first line segment is equal to the length of the first edge of the vertebra detection frame, and the first edge of the vertebra detection frame is parallel to the X axis. Referring to fig. 6, fig. 6 is a schematic view of vertebral points according to an embodiment of the present invention, in fig. 6, a first line passes through the center of a vertebral detection frame, intersects with a short side of the vertebral detection frame, and is parallel to a long side (first side) of the vertebral detection frame, a part of points located on the first line are schematically shown in fig. 6, and all points located on the first line are vertebral points in this embodiment. In other embodiments, the vertebral points may be all points located on a second line segment, which is parallel to and at a predetermined distance from the first line segment, as shown in fig. 6, which may be any of 10 pixels to 20 pixels, such as 15 pixels.

After the vertebra points in each frame of 2D medical image are determined, the vertebra information in each frame of 2D medical image is determined next. The vertebra information includes: the central point and the direction of the straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point.

In this embodiment, the direction of the straight line where the central point and the vertebra point are located can be represented by an angle between a ray formed by the central point and the vertebra point and the positive direction of the X-axis, with the central point as a starting point. Of course, in other embodiments, the center point may be used as the starting point, and the included angle between the ray formed by the center point and the vertebra point and the negative X-axis direction represents the direction of the straight line where the center point and the vertebra point are located, or the vertebra point may be used as the starting point, the included angle between the ray formed by the vertebra point and the center point and the positive X-axis direction represents the direction of the straight line where the center point and the vertebra point are located, and the included angle between the ray formed by the vertebra point and the center point and the negative X-axis direction represents the direction of the straight line where the center point and the vertebra point are located. In the actual mapping, different ways may be chosen to determine the direction of the straight line in which the center point and the vertebra point lie. Therefore, whether the center point is a positive direction or a negative direction of the X axis from the center point or the vertebra point is taken as a starting point when determining the direction of the straight line of the center point and the vertebra point should not be taken as a limitation when determining the direction of the straight line of the center point and the vertebra point.

As can be seen from the above, the size of the vertebra detection box in each frame of 2D medical image and the center coordinates of the vertebra detection box are known, so the coordinates of each vertebra point on the first line segment can also be known (the Z coordinate of the vertebra point can be determined by the position of the frame of 2D medical image). The coordinates of the center point are known, and therefore the direction of the straight line in which the center point and the vertebra point are located can be determined from the coordinates of the center point and the coordinates of the vertebra point.

After the vertebra point information in each frame of 2D medical image is determined, S14 is executed to map the vertebra point information in the frames of 2D medical images to the first coordinate system to obtain the medical image corresponding to the frames of 2D medical images. Specifically, in this embodiment, the direction of the straight line where the central point and the vertebra point are located is mapped to the first coordinate axis, and the Z coordinate of the vertebra point is mapped to the second coordinate axis. The first coordinate axis is perpendicular to the second coordinate axis. Namely, the coordinate value of the vertebra point on the first coordinate axis is determined based on the direction of the straight line where the central point and the vertebra point are located, and the coordinate value of the vertebra point on the second coordinate axis is determined based on the Z coordinate of the vertebra point, so that the vertebra point corresponding to the vertebra point can be determined on the first coordinate system. The pixel values of the vertebra points may not change before and after mapping.

The mapping process is explained below by way of a simple example.

For a plurality of vertebra points located in the vertebra detection frame on a certain frame of 2D medical image, taking a first vertebra point, a second vertebra point and a third vertebra point as an example, the direction of a straight line where the central point and the first vertebra point are located is 45 degrees, the Z coordinate of the first vertebra point is 300mm, the direction of a straight line where the central point and the second vertebra point are located is 50 degrees, the Z coordinate of the second vertebra point is 300mm, the direction of a straight line where the central point and the third vertebra point are located is 55 degrees, and the Z coordinate of the third vertebra point is 300 mm. Then 45 ° corresponding to the first vertebra point is mapped to a first coordinate axis (angular axis) of the first coordinate system, and 300mm corresponding to the first vertebra point is mapped to a second coordinate axis of the first coordinate system, so as to obtain the vertebra point corresponding to the first vertebra point in the first coordinate system. And mapping 50 degrees corresponding to the second vertebra point to a first coordinate axis of a first coordinate system, and mapping 30mm corresponding to the second vertebra point to a second coordinate axis of the first coordinate system to obtain the vertebra point corresponding to the second vertebra point in the first coordinate system. And mapping 50 degrees corresponding to the third vertebra point to a first coordinate axis of a first coordinate system, and mapping 300mm corresponding to the third vertebra point to a second coordinate axis of the first coordinate system to obtain the vertebra point corresponding to the third vertebra point in the first coordinate system. The vertebra point information in the multi-frame 2D medical images is mapped according to the method, so that the vertebra in the multi-frame 2D medical images can be mapped into a vertebra panorama.

Thus, the vertebra points in each frame of 2D medical image are correspondingly mapped in the above manner, so that a vertebra panorama corresponding to the plurality of frames of 2D medical images is obtained. As the vertebrae in the 2D medical image are not required to be mapped into the vertebrae panorama in an artificial mode, the mapping efficiency and the mapping accuracy are improved, and the diagnosis efficiency and the diagnosis accuracy of a doctor are also improved to a certain extent.

Based on the same technical concept, an embodiment of the present invention provides a medical image processing apparatus, including:

a first acquisition unit for acquiring a rib point in the 3D medical image.

A determination unit for determining a center point (x)₀，y₀)，x₀Associated with the x-coordinate, y, of the rib point₀The Y-coordinate associated with the rib points, the X-axis is perpendicular to the sagittal plane and the Y-axis is perpendicular to the coronal plane.

The implementation of the medical image processing apparatus of this embodiment can refer to the implementation of the medical image processing method described above, and is not described herein again.

Based on the same technical concept, embodiments of the present invention provide a computer device, which includes at least one processor and at least one memory, wherein the memory stores a computer program, and when the program is executed by the processor, the processor is enabled to execute the medical image processing method.

Based on the same technical concept, embodiments of the present invention provide a computer-readable storage medium, wherein instructions of the storage medium, when executed by a processor in a device, enable the device to perform the above-mentioned medical image processing method.

Example two

In the process of mapping the vertebra point in each frame of 2D medical image to the vertebra panorama, in order to improve the display effect of the mapped vertebra panorama, this embodiment provides a medical image processing method, referring to fig. 7, where fig. 7 is a flowchart illustrating a medical image processing method according to a second embodiment of the present invention, as shown in fig. 7, the medical image processing method includes:

s21: rib points in the 3D medical image are acquired.

S22: determining a center point (x)₀，y₀)，x₀Associated with the x-coordinate, y, of the rib point₀In relation to the y-coordinate of the rib point,the X-axis is perpendicular to the sagittal plane and the Y-axis is perpendicular to the coronal plane.

S23: acquiring a vertebra detection frame in any frame of 2D medical image.

S24: and translating the vertebra detection frame in each acquired frame of 2D medical image to obtain a translated vertebra detection frame corresponding to the vertebra detection frame, wherein the Y coordinates of the center of the translated vertebra detection frame in each frame of 2D medical image are the same.

S25: acquiring vertebra information in any frame of 2D medical image, wherein the vertebra information comprises: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is away from the first line segment by a preset distance, and the first line segment passes through the center of the translated vertebra detection frame of the frame of 2D medical image and is parallel to the X axis.

S26: mapping vertebra information in the multiple frames of 2D medical images to a first coordinate system to obtain medical images corresponding to the multiple frames of 2D medical images.

In this embodiment, S21 to S23 and S25 to S26 are similar to those in the first embodiment, and are not described again here. Different from the first embodiment, in the present embodiment, after the vertebra detection frame in each frame of 2D medical image is detected, the detected vertebra detection frame in each frame of 2D medical image is translated to obtain the translated vertebra detection frame in each frame of 2D medical image, and the Y coordinates of the translated vertebra detection frame in each frame of 2D medical image are the same. In this embodiment, the Y coordinate of the vertebra detection frame in each frame of the 2D medical image may be a preset value, and the preset value may be determined according to actual experience. For example, the preset value may be the value with the largest Y coordinate value in all the vertebra detection boxes. After the vertebra detection frames detected in all the frame 2D medical images are translated correspondingly, vertebra information in each frame 2D medical image is obtained based on the translated vertebra detection frames, corresponding mapping is further carried out, and a vertebra panoramic image with high mapping precision and good mapping effect can be obtained.

Based on the same technical concept, the present embodiment further provides a medical image processing apparatus, including:

a first acquisition unit for acquiring a rib point in the 3D medical image.

And the third acquisition unit is used for acquiring the vertebra detection frame in any frame of the 2D medical image.

And the translation unit is used for translating the vertebra detection frame in each acquired frame of 2D medical image to obtain a translated vertebra detection frame corresponding to the vertebra detection frame, wherein the Y coordinates of the center of the translated vertebra detection frame in each frame of 2D medical image are the same.

A fourth obtaining unit, configured to obtain vertebra information in any frame of the 2D medical image, where the vertebra information includes: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is away from the first line segment by a preset distance, and the first line segment passes through the center of the translated vertebra detection frame of the frame of 2D medical image and is parallel to the X axis.

EXAMPLE III

The present embodiment is different from the above-described embodiments in that a rib panorama is displayed at the same time as a vertebra panorama is displayed in the present embodiment.

Fig. 8 is a flowchart illustrating a medical image processing method according to a third embodiment of the present invention, and as shown in fig. 8, the medical image processing method includes:

s31: rib points in the 3D medical image are acquired.

S32: determining a center point (x)₀，y₀)，x₀Associated with the x-coordinate, y, of the rib point₀The Y-coordinate associated with the rib points, the X-axis is perpendicular to the sagittal plane and the Y-axis is perpendicular to the coronal plane.

S33: rib information in any frame of 2D medical image is acquired, wherein the rib information comprises: the center point and the direction of a straight line where each rib point in the frame of 2D medical image is located, and the Z coordinate of the rib point.

S34: acquiring a vertebra detection frame in any frame of 2D medical image.

S35: and translating the vertebra detection frame in each acquired frame of 2D medical image to obtain a translated vertebra detection frame corresponding to the vertebra detection frame, wherein the Y coordinates of the center of the translated vertebra detection frame in each frame of 2D medical image are the same.

S36: acquiring vertebra information in any frame of 2D medical image, wherein the vertebra information comprises: the central point and the direction of a straight line where each vertebra point in the frame of 2D medical image is located, and the Z coordinate of the vertebra point, the vertebra point is located on a first line segment or a second line segment which is parallel to the first line segment and is away from the first line segment by a preset distance, and the first line segment passes through the center of the translated vertebra detection frame of the frame of 2D medical image and is parallel to the X axis.

S37: mapping rib information and vertebra information in the multi-frame 2D medical images to a first coordinate system to obtain medical images corresponding to the multi-frame 2D medical images.

In this embodiment, the determination of the center point of S32, the mapping of the vertebra information in S35-S37 and the multiple frames of 2D medical images is similar to that in the first embodiment or the second embodiment, and will not be described herein again. The following mainly describes the rib mapping in each frame of 2D medical image.

As can be seen from the first embodiment, the rib points in the 3D medical image may be points (rib positioning points) obtained by detecting the rib points in the 3D medical image, in this embodiment, in order to obtain as many rib points as possible and further improve the accuracy of the rib panorama obtained by final mapping, expansion may be performed based on the rib positioning points, for example, associated positioning points associated with the rib positioning points may be used as the rib points, and in addition, points (rib contour points) on the rib contour may also be used as the rib points.

Specifically, the rib positioning points in each frame of 2D medical image may be correspondingly extended to obtain associated positioning points associated with the rib positioning points, and the associated positioning points are used as supplements to the rib points. Further, the rib contour in each frame of the 2D medical image (cross section) is detected, and a point on the detected rib contour (rib contour point) is taken as a rib point.

In this embodiment, specifically, the rib positioning point in each frame of the 2D medical image (cross-sectional image) is expanded in the following manner to obtain an associated positioning point associated with the rib positioning point. For any rib positioning point in any frame of 2D medical image, first, the center point (x) is used₀，y₀) And drawing an arc by taking the distance between the center point and the rib positioning point as a radius as a circle center, and then taking a point on the arc near the rib positioning point as an associated positioning point associated with the rib positioning point. Specifically, when an arc is drawn by taking the center point as the center of a circle and the distance between the center point and the rib positioning point as a radius, the formed arc may be symmetrical by taking the rib positioning point as the center, and the radian corresponding to the arc may be between 3 ° and 7 °, for example: the arc may correspond to an arc of 5 °. In this embodiment, all points located around the rib positioning point on the generated arc may be used as associated positioning points associated with the rib positioning point, and these associated positioning points may be used as rib points.

In order to obtain points on the rib contour, the rib contour needs to be detected first. In the present embodiment, specifically, the rib contour in any one frame of the 2D medical image (cross-sectional image) is acquired by the following steps.

Firstly: obtaining key points in the 2D medical image, wherein the key points are related to ribs. The key point may be located on the rib or near the rib, and for one rib, there may be one key point or multiple key points. Referring to fig. 9, fig. 9 is a schematic diagram of key points according to an embodiment of the present invention, where the points identified by numbers in fig. 9 are the key points, and fig. 9 schematically identifies 20 key points.

In this embodiment, the positions of the key points are obtained through a neural network, and specifically, the positions of the key points may be obtained through a 2D regression network. The 2D regression network comprises a feature extraction unit and a fully connected regression unit, and the output of the feature extraction unit is the input of the fully connected regression unit. The feature extraction unit comprises N convolution modules and N maximum pooling layers, wherein the output of each convolution module is connected with the input of each maximum pooling layer, namely the convolution modules are alternately connected with the maximum pooling layers. Each convolution module includes a plurality of convolution blocks, wherein each convolution block includes: a convolutional layer (Conv2d), a Batch Normalization layer (BN), and an activation layer, which may be a Linear rectification function (ReLU). The full-connection regression unit comprises M sequentially continuous full-connection layers, and a dropout layer with the passing rate of 0.5 can be arranged between each full-connection layer and each full-connection layer. The fully connected regression unit finally outputs coordinates of the key points, which may be pixel coordinates, for example, the abscissa and the ordinate of the key point may be the corresponding number one pixel in a preset coordinate system.

In this embodiment, multiple frames of 2D medical images (cross-sectional images) may be used as training samples, and a labeling person labels key points on each frame of 2D medical image, where the key points may be points on or near each rib, and the number of the key points may be determined according to actual requirements. And then, performing data enhancement on the training sample (such as random rotation at a certain angle, random up-down and left-right translation of 0-30 pixels, random scaling of 0.85-1.15 times, small amount of jitter on image contrast and brightness and the like), and enhancing the data amount to 10 times of the original data amount. And finally, inputting the training sample into a 2D regression network for training. During training, the loss function is calculated according to the coordinates of the marked key points and the coordinates of the key points predicted by the network, training is performed by a back propagation method, and the training optimization algorithm can adopt an SGD algorithm with momentum and step attenuation. After the 2D regression network is obtained through training, the 2D medical image may be input to the 2D regression network to obtain coordinates of key points corresponding to the ribs in the 2D medical image.

So far, the coordinates of the key points in the 2D medical image are obtained through the 2D regression network, and then the key points are connected to obtain the rib contour in the 2D medical image. Finally, rib contour points on the rib contour, which do not belong to rib positioning points and associated positioning points associated with the rib positioning points, can be used as rib points.

After obtaining the rib points in each frame of 2D medical image, S33 is executed to obtain rib information in each frame of 2D medical image, where the rib information includes: the center point and the direction of a straight line where each rib point in the frame of 2D medical image is located, and the Z coordinate of the rib point. In this embodiment, the direction of the straight line where the center point and the rib point are located may be represented by an angle between a ray formed by the center point and the rib point and the positive direction of the X-axis, with the center point as a starting point. Of course, in other embodiments, the center point may be used as the starting point, and the included angle between the ray formed by the center point and the rib point and the negative direction of the X axis represents the direction of the straight line where the center point and the rib point are located. In the actual mapping, different ways can be selected to determine the direction of the straight line where the center point and the rib point are located. Therefore, when determining the direction of the straight line of the center point and the rib point, whether the center point is used as the starting point or the rib point is used as the starting point, the included angle between the center point and the rib point and the positive direction or the negative direction of the X axis should not be taken as the definition when determining the direction of the straight line of the center point and the rib point.

From the above, when the 3D medical image is detected, the coordinates of the detected rib positioning point are known, and in each frame of the 2D medical image, the coordinates of the associated positioning point associated with the rib positioning point may also be determined according to the distance between the central point and the rib positioning point and the corresponding radian. The coordinates of each rib contour point on the rib contour are also known in each frame of the 2D medical image. Therefore, in this embodiment, the direction of the straight line where the center point and the rib point are located can be determined according to the coordinates of the center point and the coordinates of the rib point.

After determining the rib information in each frame of the 2D medical image, S34 is executed to map the rib information in the frames of the 2D medical image to the first coordinate system to obtain the medical image corresponding to the frames of the 2D medical image. Specifically, in this embodiment, the direction of the straight line where the center point and the rib point are located is mapped to the first coordinate axis, and the Z coordinate of the rib point is mapped to the second coordinate axis. The first coordinate axis is perpendicular to the second coordinate axis. That is, the coordinate value of the rib point on the first coordinate axis is determined based on the direction of the straight line where the center point and the rib point are located, and the coordinate value of the rib point on the second coordinate is determined based on the Z coordinate of the rib point, so that the rib point corresponding to the rib point can be determined on the first coordinate system. The pixel value of the rib point may not change before and after mapping.

The mapping process is explained below by way of a simple example.

If a plurality of rib points exist on a certain frame of 2D medical image, taking a first rib point, a second rib point, and a third rib point as an example, the direction of a straight line where the center point and the first rib point are located is 10 °, the Z coordinate of the first rib point is 300mm, the direction of a straight line where the center point and the second rib point are located is 13 °, the Z coordinate of the second rib point is 320mm, the direction of a straight line where the center point and the third rib point are located is 16 °, and the Z coordinate of the third rib point is 340 mm. Then 10 ° corresponding to the first rib point is mapped to a first coordinate axis (angular axis) of the first coordinate system, and 300mm corresponding to the first rib point is mapped to a second coordinate axis of the first coordinate system, so as to obtain a rib point corresponding to the first rib point in the first coordinate system. And mapping 13 degrees corresponding to the second rib point to a first coordinate axis of a first coordinate system, and mapping 320mm corresponding to the second rib point to a second coordinate axis of the first coordinate system to obtain the rib point corresponding to the second rib point in the second coordinate system. Mapping 16 degrees corresponding to the third rib point to a first coordinate axis of a first coordinate system, and mapping 340mm corresponding to the third rib point to a second coordinate axis of the first coordinate system to obtain the rib point corresponding to the third rib point in the first coordinate system. The rib information in the multi-frame 2D medical images is mapped according to the method, so that the ribs in the multi-frame 2D medical images can be mapped into a rib panorama.

It should be noted that, in the process of mapping the rib panorama, the rib points include rib positioning points, associated positioning points associated with the rib positioning points, and rib contour points. In other embodiments, when the predetermined interval is small enough and the first tangent plane and the second tangent plane are large enough, the detected rib positioning points may also be used as rib points in the 3D medical image, rib information is obtained based on the rib positioning points in each frame of the 2D medical image, and the rib panorama is mapped based on the rib information in the multi-frame 2D medical image.

Thus, the ribs and vertebrae in the multi-frame 2D medical image are mapped correspondingly through the above-mentioned S31 to S37, so as to obtain a panoramic image. Because the mapping process does not need to be carried out in a manual mode, the mapping accuracy and the mapping efficiency are improved. In addition, when the ribs in each frame of 2D medical image are mapped, the acquisition of rib points is not limited to the rib positioning points obtained by detecting the 3D medical image, the rib points are supplemented based on the rib positioning points, and the detected points on the rib contour in each frame of 2D medical image are also used as a part of the rib points to be acquired, so that the number of the obtained rib points is increased to a great extent, and the accuracy of the finally obtained panoramic image is also increased to a great extent.

a first acquisition unit for acquiring a rib point in the 3D medical image.

A fifth obtaining unit, configured to obtain rib information in any frame of the 2D medical image, where the rib information includes: the center point and the direction of a straight line where each rib point in the frame of 2D medical image is located, and the Z coordinate of the rib point.

And the third mapping unit is used for mapping rib information and vertebra information in the multi-frame 2D medical images to the first coordinate system so as to obtain the medical images corresponding to the multi-frame 2D medical images.

Based on the same technical concept, the present embodiment also provides a computer device, which includes at least one processor and at least one memory, where the memory stores a computer program, and when the program is executed by the processor, the processor is enabled to execute the medical image processing method.

Based on the same technical concept, the present embodiment also provides a computer-readable storage medium, wherein when the instructions in the storage medium are executed by a processor in the apparatus, the apparatus is enabled to execute the medical image processing method.

It should be apparent to those skilled in the art that embodiments of the present invention may be provided as a method, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.

Claims

1. A method of medical image processing, comprising:

acquiring rib points in the 3D medical image;

determining a center point (x)₀,y₀)，x₀Associated with the x-coordinate, y, of the rib point₀A Y-coordinate associated with the rib point, the X-axis being perpendicular to the sagittal plane and the Y-axis being perpendicular to the coronal plane;

2. The method of claim 1, further comprising:

3. The method of claim 2, wherein the rib site comprises: rib positioning point, associated positioning point associated with said rib positioning point, said x₀X-coordinate, y, associated with the positioning point of the rib₀The y-coordinate associated with the rib positioning point.

4. The method of claim 2, wherein the rib site comprises: rib positioning point, associated positioning point associated with the rib positioning point, rib contour point, the x₀X-coordinate, y, associated with the positioning point of the rib₀The y-coordinate associated with the rib positioning point.

5. The method of claim 2, wherein mapping the vertebra information in the plurality of frames of 2D medical images to the first coordinate system comprises:

6. The method of claim 3 or 4, further comprising obtaining rib positioning points, the obtaining rib positioning points comprising:

7. The method of claim 6, wherein the first location point is a center point or a center of gravity point of the left lung and the second location point is a center point or a center of gravity point of the right lung.

8. The method according to any one of claims 1 to 5,

the length of the first line segment is equal to the length of a first side of the vertebra detection frame, and the first side of the vertebra detection frame is parallel to the X axis.

9. A method of medical image processing, comprising:

acquiring rib points in the 3D medical image;

acquiring a vertebra detection frame in any frame of 2D medical image;

10. The method of claim 9, further comprising: